274 THE POWER OF RESISTANCE TO EXTREMES 



without these additions when one gramme-molecule is dissolved in 51,200 litres 

 of water. Similarly a solution of silver nitrate containing one gramme-molecule in 

 204,800 litres exercises the same effect as a solution of one gramme-molecule of 

 AgNO.j-f3KCN in 25,600 litres. The poisonous action is therefore decreased, 

 owing to the formation of the feebly dissociating cyanide of silver, in spite of the 

 addition of potassium cyanide, which is a strong poison, though not as strong as 

 is silver nitrate. Kronig and Paul have, however, found that the presence of a little 

 alcohol increases the poisonous action of mercuric chloride and of silver nitrate, 

 hut not of phenol, either because of its action on metabolism, or because it 

 increases the rapidity of absorption. Strong alcohol, on the other hand, greatly 

 decreases the poisonous action in plants able to withstand the consequent with- 

 drawal of water. 



It follows from the above that the disinfecting power of a poison depends 

 upon a variety of factors, and is not the direct measure of the quantity of the 

 poisonous base or acid present, as has been frequently incorrectly assumed l . 



Non-electrolytes are also poisonous, and by chemical decomposition the 

 poisonous properties may be altered in the same way as when an electrolyte 

 undergoes dissociation. The form in which a compound is presented is of great 

 importance even in a dissociating electrolyte. Thus in the form of ferrocyanogen 

 iron has a dissimilar action to that exerted by it as a base combined with an acid. 

 Complex metallic ions appear to be less poisonous than simple ones, and hence 

 in general the poisonous action of metallic salts appears to be increased by 

 dissociation. It is, however, possible that with some substances the reverse may 

 be the case, just as the products of the decomposition of a non-electrolyte may be 

 more or less poisonous than the original substance. Further, an innocuous compound 

 of a poisonous metal, which does not dissociate in water, will act as a poison if the 

 protoplast withdraws the poisonous metallic ions from the compound. 



1 Cf. Fliigge, Mikroorganismen, 1896, 3. Aufl., Bd. I, p. 453. 



